Mercurial > repos > davidvanzessen > sff_extract_demultiplex
diff trim.py @ 3:79be0752711d draft
Uploaded
author | davidvanzessen |
---|---|
date | Tue, 01 Jul 2014 08:27:57 -0400 |
parents | |
children | 8e3d95d7f342 |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/trim.py Tue Jul 01 08:27:57 2014 -0400 @@ -0,0 +1,35 @@ +import argparse + +#docs.python.org/dev/library/argparse.html +parser = argparse.ArgumentParser() +parser.add_argument("--input", help="Input folder with files") +parser.add_argument("--output", help="Output file") +parser.add_argument("--start", help="How many nucleotides to trim from the start", type=int) +parser.add_argument("--end", help="How many nucleotides to trim from the end", type=int) + +args = parser.parse_args() +start = int(args.start) +end = int(args.end) + +if end <= 0: + import shutil + shutil.copy(args.input, args.output) + import sys + sys.exit() + +currentSeq = "" +currentId = "" + +with open(args.input, 'r') as i: + with open(args.output, 'w') as o: + for line in i.readlines(): + if line[0] is ">": + if currentSeq is not "" or currentId is not "": + o.write(currentId) + o.write(currentSeq[start:-end] + "\n") + currentId = line + currentSeq = "" + else: + currentSeq += line.rstrip() + o.write(currentId) + o.write(currentSeq.rstrip()[start:-end] + "\n")